416. Partition Equal Subset Sum

Problem Description

The given problem asks us to determine if we can split an array of integers, nums, into two subsets such that the sum of the elements in both subsets is the same. This is essentially asking if nums can be partitioned into two subsets of equal sum. If such a partition is possible, we should return true, otherwise, we return false.

To understand this problem better, imagine you have a set of blocks with different weights, and you want to see if you can divide them into two groups that weigh the same. If it can be done, then each group represents a subset with an equal sum.

Intuition

The solution to this problem is based on the concept of dynamic programming, particularly the 0/1 Knapsack problem, where we aim to find a subset of numbers that can sum up to a specific target, in this case, half the sum of all elements in nums.

The intuition behind this solution is:

First, we calculate the sum of all elements in the array. If the sum is an odd number, it's impossible to partition the array into two subsets with an equal sum, so we immediately return false.
If the sum is even, our target becomes half of the total sum, and we set up an array f of boolean values that represents if this sum can be reached using a combination of the numbers we’ve seen so far. f is initialized with a size equal to the target plus one (m + 1), with the first value True (since we can always reach a sum of 0) and the rest False.
We iterate over each number in our array nums. For each number, we update our f array from right to left, starting at our target m and going down to the value of the number x. We do this backward to ensure that each number is only considered once. At each position j, we update f[j] by checking if f[j] was previously true or if f[j-x] was true. The latter means that if we already could sum up to j-x, then by adding x, we can now also sum up to j.
At the end of this process, f[m] tells us whether we've found a subset of elements that sum up to m, which would be half the sum of the entire array. If f[m] is true, we have our partition and return true, otherwise, we return false.

Learn more about Dynamic Programming patterns.

Solution Approach

The solution implements a classic dynamic programming approach to solve the subset sum problem, which is a variation of the 0/1 Knapsack problem. Here's a step-by-step guide to understanding the algorithm:

Calculate the Sum and Determine Feasibility: We begin by finding the sum of all elements in the array using sum(nums). We divide this sum by 2 using the divmod function, which gives us the quotient m and the remainder mod. If mod is not zero, the sum is odd, and we cannot partition an odd sum into two equal even halves, so we return false.
Dynamic Programming Array Setup: Next, we set up an array f with m + 1 boolean elements, which will help us track which sums can be achieved from subsets of the array. We initialize f[0] to True because a zero sum is always possible (the empty subset), and the rest to False.
Iterate and Update the DP Array: For each number x in nums, we iterate over the array f from m down to x. We do this in reverse order to ensure that each element contributes only once to each sum.
Update the DP Array: For each position j in f, we check if f[j] was already True (sum j was already achievable) or if f[j - x] was True. If f[j - x] was True, it means there was a subset of previous elements that added up to j - x. By including the current element x, we can now reach the sum j, so we set f[j] to True.
Return the Result: Finally, we return the value of f[m]. This value tells us whether there is a subset of elements from nums that adds up to m, which would be half of the total sum. If f[m] is True, it means we can partition the array into two subsets with an equal sum, and we return true; otherwise, we return false.

The pattern used in this algorithm leverages the properties of boolean arithmetic wherein True represents 1 and False represents 0. The statement f[j] = f[j] or f[j - x] is an efficient way to update our boolean array because it captures both conditions for setting f[j] to True: either it's already True, or f[j - x] is True and we just add x to reach the required sum j.

By re-using the array f in each iteration and only considering each number once, we keep our space complexity to O(sum/2), which is much more efficient than trying to store all possible subset sums up to the total sum of the array.

Ready to land your dream job?

Unlock your dream job with a 2-minute evaluator for a personalized learning plan!

Start Evaluator

Example Walkthrough

Let's walk through an example to illustrate the solution approach. Consider an array nums with the following elements: [1, 5, 11, 5].

Calculate the Sum and Determine Feasibility:
- Compute the sum of the elements: 1 + 5 + 11 + 5 = 22.
- Use divmod to check if the sum is even or odd: divmod(22, 2) gives us (11, 0).
- Since the remainder is 0, the sum is even, and proceeding is feasible.
Dynamic Programming Array Setup:
- Our target sum m is 22 / 2 = 11.
- Initialize f with dimensions [12] (m + 1) and set f[0] to True.
Iterate and Update the DP Array:
- Start iterating over the array nums: [1, 5, 11, 5].
Update the DP Array:
- For x = 1 (first element), update f from 11 down to 1. Since f[0] is True, set f[1] to True.
- For x = 5 (second element), update f from 11 down to 5. Now f[5], f[6], f[7], f[8], f[9], and f[11] become True.
- For x = 11 (third element), since f[0] is True, set f[11] to True. However, f[11] is already True from the previous step.
- Lastly, for x = 5 (fourth element), update f again similarly to when x was 5 before.
Return the Result:
- After the final iteration, we check the value of f[11], which is True.
- This indicates that there is a subset with a sum of 11, which is half of the total sum.

Therefore, the array [1, 5, 11, 5] can be partitioned into two subsets with equal sum, and we return true.

Solution Implementation

1class Solution:
2    def can_partition(self, nums: List[int]) -> bool:
3        # Compute the total sum of the nums array and divide by 2 (partition sum)
4        total_sum, remainder = divmod(sum(nums), 2)
5      
6        # If the sum of nums is odd, we cannot partition it into two equal subsets
7        if remainder:
8            return False
9      
10        # Initialize a boolean array that will keep track of possible sums
11        can_partition = [True] + [False] * total_sum
12      
13        # Loop through each number in the nums array
14        for num in nums:
15            # Check each possible sum in reverse (to avoid using the same number twice)
16            for j in range(total_sum, num - 1, -1):
17                # Update the can_partition array
18                # True if the number itself can form the sum
19                # or if the sum can be formed by adding the number to a previously possible sum
20                can_partition[j] = can_partition[j] or can_partition[j - num]
21      
22        # The last element in the can_partition array indicates if we can partition
23        # nums into two equal subsets
24        return can_partition[total_sum]
25

1class Solution {
2    public boolean canPartition(int[] nums) {
3        // Calculate the sum of all array elements
4        int sum = 0;
5        for (int num : nums) {
6            sum += num;
7        }
8      
9        // If the sum is odd, it's not possible to partition the array into two subsets of equal sum
10        if (sum % 2 != 0) {
11            return false;
12        }
13      
14        // Target sum for each subset is half of the total sum
15        int targetSum = sum / 2;
16      
17        // Create a boolean array to store the subset sums achievable up to the targetSum
18        boolean[] subsetSums = new boolean[targetSum + 1];
19      
20        // There's always one subset with sum 0, the empty set
21        subsetSums[0] = true;
22      
23        // Check each number in the given array
24        for (int num : nums) {
25            // Traverse the subsetSums array in reverse to avoid using an element multiple times
26            for (int j = targetSum; j >= num; j--) {
27                // Update the subset sums that are achievable
28                // If j-num is achievable, set j as achievable (because we're adding num to the subset)
29                subsetSums[j] = subsetSums[j] || subsetSums[j - num];
30            }
31        }
32      
33        // The result is whether the targetSum is achievable as a subset sum
34        return subsetSums[targetSum];
35    }
36}
37

1#include <numeric>
2#include <vector>
3#include <cstring>
4
5class Solution {
6public:
7    // Function to determine if the input array can be partitioned into two subsets of equal sum
8    bool canPartition(vector<int>& nums) {
9        // Calculate the sum of elements in the nums array
10        int totalSum = accumulate(nums.begin(), nums.end(), 0);
11
12        // If the total sum is odd, it's not possible to divide it into two equal parts
13        if (totalSum % 2 == 1) {
14            return false;
15        }
16
17        // Target sum for each partition
18        int targetSum = totalSum >> 1;
19
20        // Create a dynamic programming array to keep track of possible sums
21        bool dp[targetSum + 1];
22
23        // Initialize the dynamic programming array to false
24        memset(dp, false, sizeof(dp));
25
26        // The sum of 0 is always achievable (by selecting no elements)
27        dp[0] = true;
28
29        // Iterate through the numbers in the array
30        for (int num : nums) {
31            // Check each possible sum in reverse to avoid using a number twice
32            for (int j = targetSum; j >= num; --j) {
33                // Update the dp array: dp[j] will be true if dp[j - num] was true
34                // This means that current number 'num' can add up to 'j' using the previous numbers
35                dp[j] = dp[j] || dp[j - num];
36            }
37        }
38
39        // The result is whether it's possible to achieve the targetSum using the array elements
40        return dp[targetSum];
41    }
42};
43

1function canPartition(nums: number[]): boolean {
2    // Calculate the sum of all elements in the array
3    const totalSum = nums.reduce((accumulator, currentValue) => accumulator + currentValue, 0);
4
5    // If the total sum is odd, it's not possible to partition the array into two subsets with an equal sum
6    if (totalSum % 2 !== 0) {
7        return false;
8    }
9
10    // Target sum is half of the total sum
11    const targetSum = totalSum >> 1;
12
13    // Initialize a boolean array to keep track of possible subset sums
14    const possibleSums = Array(targetSum + 1).fill(false);
15    // Always possible to pick a subset with sum 0 (empty subset)
16    possibleSums[0] = true;
17
18    // Iterate through all numbers in the given array
19    for (const num of nums) {
20        // Iterate backwards through possibleSums array to check if current number can contribute to the targetSum
21        for (let j = targetSum; j >= num; --j) {
22            // Update possibleSums array to reflect the new subset sum that can be formed
23            possibleSums[j] = possibleSums[j] || possibleSums[j - num];
24        }
25    }
26
27    // Return whether a subset with the targetSum is possible
28    return possibleSums[targetSum];
29}
30

Time and Space Complexity

The code is designed to solve the Partition Equal Subset Sum problem which is to determine if the given set of numbers can be partitioned into two subsets such that the sum of elements in both subsets is the same.

Time Complexity

The time complexity is O(n * m) where n is the number of elements in nums and m is half the sum of all elements in nums if the sum is even. This complexity arises from the double loop structure: an outer loop iterating over each number x in nums, and an inner loop iterating backwards from m to x. The inner loop runs at most m iterations (representing the possible sums up to half the total sum), and this is done for each of the n numbers.

Space Complexity

The space complexity is O(m) where m is half the sum of all elements in nums (if the sum is even). This is due to the array f, which stores Boolean values indicating whether a certain sum can be reached with the current subset of numbers. The array f has a length of m + 1, with m being the target sum (the zero is included to represent the empty subset).

Learn more about how to find time and space complexity quickly using problem constraints.

Discover Your Strengths and Weaknesses: Take Our 2-Minute Quiz to Tailor Your Study Plan:

Consider the classic dynamic programming of longest increasing subsequence:

Find the length of the longest subsequence of a given sequence such that all elements of the subsequence are sorted in increasing order.

For example, the length of LIS for [50, 3, 10, 7, 40, 80] is 4 and LIS is [3, 7, 40, 80].

What is the recurrence relation?

dp[i] = dp[i] + 1

dp[i] = dp[i] + dp[i - 1]

dp[i] = (dp[i] + dp[j] + 1) for j in 0 to i

dp[i] = max(dp[i], dp[j] + 1) for j in 0 to i

I don't know